Tax Efficient Withdrawals: How to Pay Less in Retirement

Most retirement planning focuses on accumulation: how much to save, where to invest, and when you have enough to stop working. Far less attention goes to the distribution phase — specifically, which accounts to draw from first, how much, and why the order matters so much.

A retiree with $2M split across taxable, traditional, and Roth accounts has many options for generating $80,000/year in income. But not all options are equal. The wrong order can push income into higher brackets, trigger ACA subsidy clawbacks, accelerate RMDs, or waste the 0% capital gains rate. The right order avoids all of those — legally and without complexity.

The optimal withdrawal order (general framework)

Bracket filling: the core skill

The concept of "bracket filling" is the most important technique in tax-efficient retirement withdrawal. Each year, you deliberately draw enough from traditional (pre-tax) accounts to fill your current tax bracket — converting or withdrawing up to the top of the 12% bracket, for example — and then switch to tax-free sources for the rest.

In 2026, for a married couple filing jointly, the 12% bracket runs from $24,800 to $100,800 of taxable income. With the standard deduction of $32,200, that means you can have up to $133,000 in total income before your withdrawals start hitting the 22% bracket.

The couple draws $77,000 from traditional sources — enough to use the full standard deduction, fill the entire 10% bracket, and fill part of the 12% bracket (which has $76,000 of total room, from $24,800 to $100,800, so this fills only a fraction of it) — paying approximately $4,880 in federal income tax (a 6.3% effective rate on that withdrawal, or 5.4% against their full $90,000 of spending). They then draw the remaining $13,000 of their $90,000 spending from Roth contributions or taxable basis — completely tax-free.

Roth conversions: filling the bracket proactively

In early retirement years when your income is especially low, you may not need to withdraw much from any account — your spending is met by taxable accounts and Roth contributions. But you can still fill your tax bracket proactively by doing Roth conversions.

If your taxable income before conversions is only $40,000 (already $15,200 into the 12% bracket, which spans $24,800–$100,800 for a married couple in 2026), you could convert up to roughly $60,800 more of traditional IRA to Roth and still stay within the 12% bracket, bringing total taxable income to $100,800 — right at the 22% threshold. You pay 10–12% on the converted amount now, rather than 22–32% when RMDs force those withdrawals later.

This proactive bracket filling during low-income years is one of the highest-value moves in FIRE retirement planning — it permanently moves money from a taxable bucket to a tax-free one at the lowest possible rate.

Real example: Kevin and Mia, retired at 52

Kevin and Mia have $1.8M total: $900k traditional IRA, $500k Roth IRA ($200k contributions, $300k earnings), $400k taxable brokerage. They spend $85,000/year. Here is their Year 1 withdrawal plan:

Source	Amount	MAGI impact	Tax
Taxable account (cost basis return)	$25,000	$0	$0
Roth IRA contributions	$20,000	$0	$0
Traditional IRA withdrawal (bracket fill)	$40,000	$40,000	~$780 (2% effective)
Total spending	$85,000	$40,000 MAGI	~$780 total

With a $32,200 standard deduction, only $7,800 of that $40,000 traditional withdrawal is taxable, and it falls entirely in the 10% bracket — hence the ~$780 tax bill on $85,000 of spending, a 0.9% effective rate. Compare this to simply drawing all $85,000 from the traditional IRA with no bracket-fill strategy: taxable income of $52,800 pushes $28,000 into the 12% bracket, for a tax bill of approximately $5,840 (6.9% effective) — a saving of roughly $5,060 per year from bracket-filling alone. Over 20 years, that difference compounds to over $100,000 in preserved wealth.

On top of covering their spending, Kevin and Mia also convert an additional $40,000 from traditional IRA to Roth this year — moving further into the 12% bracket, though well short of its $100,800 ceiling — paying approximately $4,460 in additional tax on the conversion. This isn't required to cover their spending; it's a proactive move to shrink Kevin and Mia's future RMDs while their bracket is still cheap. Including the conversion, their total tax paid this year is about $5,240 (a 6.2% effective rate against their $85,000 of spending, though the conversion itself is a separate, forward-looking transaction rather than a cost of this year's spending).

Five withdrawal-order mistakes that cost real money

Even retirees who understand the theoretical order above make predictable mistakes in practice. None of these require advanced tax knowledge to avoid — they just require a plan you actually follow, instead of reacting account-by-account as bills come due.

• Draining Roth first because it "feels free." Roth withdrawals don't show up as taxable income, which makes them tempting to reach for whenever a large expense appears. But every dollar pulled from Roth early is a dollar that stops compounding tax-free for the rest of your life. A retiree who spends down $200,000 of Roth balance in the first five years of retirement, instead of stretching it across three decades, gives up the single most valuable dollar in the portfolio for short-term convenience.
• Ignoring MAGI-based programs beyond your tax bracket. Your marginal tax bracket isn't the only thing that responds to a large traditional withdrawal. ACA premium tax credits before 65, and Medicare's Income-Related Monthly Adjustment Amount (IRMAA) surcharge on Part B and Part D premiums after 65, both key off MAGI — not taxable income. A withdrawal that looks cheap on a tax-bracket basis can still trigger a subsidy cut or a premium surcharge that shows up later.
• Skipping the 0% capital gains window. Long-term capital gains stack on top of ordinary income for bracket purposes, and for a retiree whose ordinary income stays within the 12% bracket, the portion of gains that falls in that space is taxed at 0% federally. Many retirees never realize gains deliberately — they just sell what they need and pay whatever rate applies. Filling the rest of the 0% bracket with long-term gains each year, rather than letting unrealized gains pile up, effectively resets cost basis for free.
• Withdrawing round numbers instead of the exact bracket-fill amount. Pulling a flat $50,000 or $60,000 because it's a clean number, rather than calculating the exact dollar figure that fills the target bracket, routinely leaves money on the table — either by under-filling a cheap bracket or spilling a few thousand dollars into a more expensive one.
• Forgetting that RMDs eventually override your plan. Required Minimum Distributions starting at 73 aren't optional, and they're calculated off the traditional balance on December 31 of the prior year — not off what you'd prefer to withdraw. A traditional balance left to compound untouched for decades produces RMDs that can push a retiree into a much higher bracket than any voluntary bracket-filling done in the 50s and 60s. This is the core argument for doing Roth conversions before RMDs force the issue.

How ACA subsidies change the calculus before 65

For retirees under 65, before Medicare eligibility begins, health insurance is typically purchased through the ACA marketplace, where the premium tax credit is calculated based on Modified Adjusted Gross Income relative to the federal poverty line. This is where the withdrawal order can matter even more than the tax bracket itself.

A large traditional IRA withdrawal in a single year raises MAGI for that year, which can meaningfully reduce the monthly premium tax credit — sometimes by more than the income tax saved by filling the bracket in the first place. A retiree drawing primarily from taxable basis and Roth contributions during the ACA years, and reserving traditional withdrawals and conversions for the years after Medicare begins at 65, often comes out ahead even if it means filling fewer brackets early. The math has to be run year by year against the actual subsidy schedule, not assumed from the tax bracket alone.

Here's the mechanism in illustrative terms, without relying on a specific year's poverty-line table (which shifts annually and varies by household size and state): ACA subsidies phase down gradually as MAGI rises through the relevant income bands, so the effective loss isn't a cliff at one dollar amount — it's a gradual reduction spread across a range. A retiree who adds $30,000 of traditional withdrawal MAGI in a single year, when a $10,000 taxable/Roth-only withdrawal would have covered the same spending, isn't just paying ordinary income tax on the extra $20,000 — they're also giving up some slice of subsidy across that same range. Stacking a lost subsidy dollar on top of a marginal tax dollar can push the effective marginal cost of a traditional withdrawal well above the 10–12% bracket rate that a tax-only view would suggest. This is exactly why the order matters more than the bracket in the ACA years specifically.

State taxes: the withdrawal order isn't just federal

The withdrawal order strategy in this article is built around federal tax rules, but state tax treatment varies enormously and can change the optimal sequence for a specific retiree. Several states — including Florida, Texas, Nevada, Washington, Wyoming, South Dakota, Alaska, and Tennessee — have no state income tax at all, which means the withdrawal order above applies with no state-level modification.

In states that do tax income, retirement account withdrawals are usually taxed as ordinary income, the same way they're treated federally. A number of states go further and either exempt some or all retirement account withdrawals from state tax, or offer an age-based deduction that shields a portion of retirement income once a taxpayer reaches a certain age — worth checking specifically for your state, since these provisions change the relative appeal of traditional versus Roth withdrawals at the state level.

Capital gains are the area with the biggest state-level surprise. Unlike the federal system, which taxes long-term capital gains at a lower rate than ordinary income, most states with an income tax simply tax capital gains as ordinary income — there's no state-level 0% bracket to harvest into. That doesn't eliminate the federal 0% harvesting opportunity discussed above, but it does mean the total effective tax rate on realized gains is higher than the federal-only numbers in this article suggest, for retirees in high-tax states.

For retirees with genuine flexibility about where they live — a meaningful share of the FIRE community — the state tax dimension of withdrawal order can matter more than any single federal optimization described here. Relocating from a high-tax state to a no-tax state during the exact years of large Roth conversions can be worth tens of thousands of dollars, entirely separate from the federal bracket-filling math.

A second example: single filer retiring at 45

The Kevin and Mia example above is a married couple. The math changes meaningfully for a single filer, because single-filer tax brackets are roughly half the width of married-filing-jointly brackets at the same marginal rate. Here's a full example for a single retiree.

Elena retires at 45 with $1.1M: $550,000 in a traditional IRA, $350,000 in a taxable brokerage account, and $200,000 in a Roth IRA ($120,000 of contributions, $80,000 of earnings). She spends $48,000/year.

Source	Amount	MAGI impact	Tax
Taxable account (cost basis return)	$12,000	$0	$0
Roth IRA contributions	$8,000	$0	$0
Traditional IRA withdrawal (bracket fill)	$28,000	$28,000	~$1,190 (2.5% effective)
Total spending	$48,000	$28,000 MAGI	~$1,190 total

The 2026 single-filer standard deduction is $16,100 (exactly half the $32,200 married figure used above), and the 10% bracket runs from $0 to $12,400 of taxable income. Elena's $28,000 traditional withdrawal produces taxable income of $11,900 after the standard deduction — small enough to fall entirely within the 10% bracket, for a total federal tax of about $1,190. That's an effective rate of just 2.5% on her full $48,000 of annual spending.

Elena has room to do more. The 12% bracket for a single filer runs from $12,400 to $50,400 of taxable income, meaning she could convert an additional $38,500 from traditional to Roth this year — filling the rest of the 12% bracket almost exactly — paying about $4,610 in additional tax now to permanently move that money out of future RMD territory. Whether that's worth doing depends on her expected future bracket, exactly as it did for Kevin and Mia.

How the order shifts as you get older

The framework in this article isn't static. The optimal withdrawal order for a 45-year-old FIRE retiree with three account types and no RMDs looks different from the same retiree's order at 68, and different again once RMDs begin at 73.

In the years before RMDs start, the highest-value move is usually converting traditional dollars to Roth during the lowest-income years of retirement — filling cheap brackets with conversions rather than spending, since spending is already covered by taxable and Roth contribution withdrawals. This shrinks the traditional balance that will eventually generate a mandatory RMD, and it's the single biggest lever most early retirees have to control their own future tax bracket.

There's also a bracket-cliff worth planning for explicitly: what happens to a surviving spouse. If Kevin and Mia's household shifts to Mia alone, filing as a single taxpayer the year after Kevin's death, income that stayed comfortably within the married 12% bracket (which runs up to $100,800 of taxable income) could spill into the single-filer 22% bracket, since the single 12% ceiling is only $50,400 of taxable income — barely half of the married threshold. Couples with a meaningful age or health gap sometimes deliberately accelerate Roth conversions earlier — while both spouses are alive and the married bracket is available — specifically to reduce the surviving spouse's eventual tax bill.

Elena, the single filer from the example above, also has a longer runway than most people realize before RMDs become relevant. With 28 years between her retirement at 45 and RMD age at 73, and a starting traditional balance of $550,000, even modest annual conversions of $15,000–$20,000 during her low-income years could move a large share of that balance into Roth well before it's ever forced out as a mandatory distribution. The earlier this starts, the smaller each individual conversion needs to be to reach the same end state, which keeps her inside cheap brackets the entire way instead of needing a handful of large, expensive conversion years later on.

Downturns and the withdrawal order: don't sell at the bottom

Everything above assumes normal market conditions. The withdrawal order needs a temporary override during a significant market downturn, because selling depreciated shares to generate income locks in losses that a recovering market would otherwise have erased.

This is where holding a cash or short-term bond buffer inside the taxable account earns its keep: in a down year, draw spending needs from that buffer (or from Roth contributions, which are already tax-free and don't require selling appreciated shares at a loss) instead of from accounts holding depressed equity positions. Resume the normal bracket-filling order once the portfolio recovers. The specific size of that buffer, and how to structure it, is really a separate topic — the short version is that the withdrawal order in this article describes the steady-state plan, and a market downturn is exactly the scenario where deviating from it temporarily protects the portfolio.

Building your own withdrawal order: a step-by-step checklist

Turning the framework above into a plan you actually follow comes down to a handful of concrete steps, repeated every year:

• List every account and its type. Taxable brokerage, traditional IRA/401(k), Roth IRA/401(k) — note the balance and, for Roth, how much is contributions versus earnings.
• Estimate this year's spending need. Include one-time items (a new car, a big trip) separately from baseline recurring spending, since one-time spending is a good candidate for taxable basis withdrawals specifically.
• Calculate your target bracket ceiling. Add your standard deduction to the width of the brackets you're willing to fill, using the current year's figures — these change annually with inflation adjustments.
• Check for MAGI-sensitive programs. Before 65, check ACA subsidy phase-outs; after 65, check IRMAA thresholds. Both can change the optimal traditional withdrawal amount independent of the tax bracket math.
• Fill the order: RMDs (if applicable), then taxable basis, then traditional up to your target, then Roth contributions, then Roth earnings last. Convert any remaining room in your target bracket from traditional to Roth if spending doesn't use it all.
• Revisit annually. Balances, brackets, and spending needs all shift every year — a withdrawal order calculated once at retirement and never revisited slowly drifts away from optimal.

None of this requires exotic software or a paid advisor to execute in a given year, though a fee-only CPA is worth the cost the first time you set up the plan, simply to confirm the current year's bracket thresholds, deduction amounts, and any state-specific wrinkles that apply to your situation. Once the framework is set, most retirees can run the annual calculation themselves in under an hour — it's the same six steps, with updated numbers, every single year.